An LLM is stable if it reaches the same conclusion when asked the identical question multiple times. We find leading LLMs like gpt-4o, claude-3.5, and gemini-1.5 are unstable when providing answers to hard legal questions, even when made as deterministic as possible by setting temperature to 0. We curate and release a novel dataset of 500 legal questions distilled from real cases, involving two parties, with facts, competing legal arguments, and the question of which party should prevail. When provided the exact same question, we observe that LLMs sometimes say one party should win, while other times saying the other party should win. This instability has implications for the increasing numbers of legal AI products, legal processes, and lawyers relying on these LLMs.

It began with an anonymous email. "I'm genuinely so, so sorry to reach out to you," it read. Beneath the words were three links to an internet forum. "Huge trigger warning … They contain lewd photoshopped images of you." Jodie (not her real name) froze.

In the context of unprecedented U.S. Department of Defense (DoD) budgets, this paper examines the recent history of DoD funding for academic research in algorithmically based warfighting. We draw from a corpus of DoD grant solicitations from 2007 to 2023, focusing on those addressed to researchers in the field of artificial intelligence (AI). Considering the implications of DoD funding for academic research, the paper proceeds through three analytic sections. In the first, we offer a critical examination of the distinction between basic and applied research, showing how funding calls framed as basic research nonetheless enlist researchers in a war fighting agenda. In the second, we offer a diachronic analysis of the corpus, showing how a 'one small problem' caveat, in which affirmation of progress in military technologies is qualified by acknowledgement of outstanding problems, becomes justification for additional investments in research. We close with an analysis of DoD aspirations based on a subset of Defense Advanced Research Projects Agency (DARPA) grant solicitations for the use of AI in battlefield applications. Taken together, we argue that grant solicitations work as a vehicle for the mutual enlistment of DoD funding agencies and the academic AI research community in setting research agendas. The trope of basic research in this context offers shelter from significant moral questions that military applications of one's research would raise, by obscuring the connections that implicate researchers in U.S. militarism.

AI is becoming a part of our daily lives, from approving loans to diagnosing diseases. AI model outputs are used to make increasingly important decisions, based on smart algorithms and data. But if we can't understand these decisions, how can we trust them? One approach to making AI decisions more understandable is to use models that are inherently interpretable. These are models that are designed in such a way that consumers of the model outputs can infer the model's behaviour by reading the parameters of the model. Popular inherently interpretable models include Decision Trees and Linear Regression.

The old approaches to overseeing and supporting business processes are going through a change in outlook. Troublesome innovations - like canny computerization (RPA AI) - are helping boss experience officials (CXOs) re-develop their business tasks by getting enhancements. The administrative center offers crucial help and organization to the business and can assist make administration separation with business capacities like IT, HR, and money. Advanced smart CFOs and CIOs across the globe understand that endeavors to change client confronting frameworks and cycles are restricted without similarly powerful and coordinated administrative center tasks. A study discovered that 60% of client disappointment sources began in the administrative center.

This March, IES Director Mark Schneider released a blog in which he discussed exploring a partnership with the National Science Foundation (NSF) to encourage scientists with expertise in AI and related fields to address the important post-pandemic need for accelerating learning. IES is now excited to announce our resulting participation in NSF's National Artificial Intelligence (AI) Research Institutes--Accelerating Research, Transforming Society, and Growing the American Workforce solicitation. In this blog, we describe this new funding opportunity, provide examples of existing NCSER-funded research in this area, and highlight the potential for such research to further improve outcomes for learners with disabilities. With funding from the American Rescue Plan, NCSER plans to support research under Theme 6, Track B: AI-Augmented Learning for Individuals with Disabilities. Proposals must discuss how the work will respond to the needs of learners with or at risk for a disability in an area where the COVID-19 pandemic has further widened existing gaps and/or resulted in decreased access and opportunities for students with disabilities to learn and receive support services.

The National Science Foundation officially extended the reach of its National Artificial Intelligence Research Institutes across more of the United States. On the heels of funding seven institutes in 2020, the agency last week unveiled its establishment of 11 new ones--where officials will strategically pursue AI research in complex realms like augmented learning, cybersecurity, precision agriculture and more. "The expertise of the researchers engaged in the AI Research Institutes spans a wide range of disciplines, providing an integrated effort to tackle the challenges society faces, drawing upon both foundational and use-inspired research," Director of NSF's Robust Intelligence Program Rebecca Hwa told Nextgov Tuesday. "NSF has long been able to bring together numerous fields of scientific inquiry, and in this program that includes such disciplines as computer and information science and engineering, cognitive science and psychology, economics and game theory, engineering and control theory, ethics, linguistics, mathematics, and philosophy--and that has positioned us to lead in efforts to expand the frontiers of AI." In all, the 18 institutes NSF is investing in so far underpin research spanning 40 U.S. states and the District of Columbia, Hwa confirmed.

Along with essential nutrients and trace elements, vegetables provide raw materials for the food processing industry. Despite this, plant diseases and unfavorable weather patterns continue to threaten the delicate balance between vegetable production and consumption. It is critical to utilize machine learning (ML) in this setting because it provides context for decision-making related to breeding goals. Cutting-edge technologies for crop genome sequencing and phenotyping, combined with advances in computer science, are currently fueling a revolution in vegetable science and technology. Additionally, various ML techniques such as prediction, classification, and clustering are frequently used to forecast vegetable crop production in the field. In the vegetable seed industry, machine learning algorithms are used to assess seed quality before germination and have the potential to improve vegetable production with desired features significantly; whereas, in plant disease detection and management, the ML approaches can improve decision-support systems that assist in converting massive amounts of data into valuable recommendations. On similar lines, in vegetable breeding, ML approaches are helpful in predicting treatment results, such as what will happen if a gene is silenced. Furthermore, ML approaches can be a saviour to insufficient coverage and noisy data generated using various omics platforms. This article examines ML models in the field of vegetable sciences, which encompasses breeding, biotechnology, and genome sequencing.

The Pentagon's Joint Artificial Intelligence Center is recruiting businesses to help prepare military data for use with AI. The solicitation released March 31 is a sign of the AI office's shifting role from product developer to provider of AI readiness services for Defense Department components. The basic ordering agreement would allow those components and federal partners to issue task orders for the work to get data in shape for artificial intelligence -- that could include everything from capturing data to sorting it for storage to modeling how employees will use it with AI to get better insights. The Data Readiness for Artificial Intelligence Development (DRAID) Services ordering agreement will "help the DoD and Government users prepare data for use in AI applications by providing an easily accessible path to access the cutting-edge commercial services needed to meet the complex technical challenges involved in preparing data for AI," the solicitation read. "The services addressed by the DRAID span the entire AI data preparation lifecycle, from data ingestion, through labeling, right up to before model training begins," an April 1 blog post from the JAIC states.

TYSONS, Va.--(BUSINESS WIRE)--QOMPLX, a leader in cloud-native risk analytics, has entered into a definitive agreement to acquire RPC Tyche LLP ("Tyche"), a rapidly growing insurance software modeling and consulting firm based in London, Cambridge, Paris and Chicago. Tyche bolsters QOMPLX's insurance analytics offerings, and the combined business will offer more comprehensive insurance underwriting, pricing, risk modeling, capital modeling, and reserving functionality. It is an exceptional software business that combines innovative technology with actuarial expertise to help reduce the time and costs that insurers, reinsurers and intermediaries face in producing actionable data feeding today's commercial and regulatory decision-making. Tyche and QOMPLX's combined team are building the insurance data factory of the future with superior capabilities for data integration, transformation, analysis, and contextualization for corporations, employees, and consumers. Tyche's core modeling platform focuses on the complex challenges facing insurers: pricing risks, modeling and reserving capital, and improving efficiency.